DASH

News and Events

Louhi 2014: Special Issue on Health Text Mining and Information Analysis now published in BMC Medical Informatics and Decision Making

Health Informatics, Journal, Publication, SYSLAB

We are happy to announce the publication of the Louhi 2014 Special Issue on Health Text Mining and Information Analysis, published in BMC Medical Informatics and Decision Making, Volume 15, Supplement 2:

http://www.biomedcentral.com/bmcmedinformdecismak/supplements/15/S2

This special issue is edited by myself, Martin Duneld, Aron Henriksson, Maria Kvist, Maria Skeppstedt and Hercules Dalianis, and features five new contributions following the 2014 Louhi workshop held in conjunction with EACL in Gothenburg, as well as an introduction article.

Louhi 2014: Special issue on health text mining and information analysis
Sumithra Velupillai, Martin Duneld, Aron Henriksson, Maria Kvist, Maria Skeppstedt, Hercules Dalianis
BMC Medical Informatics and Decision Making 2015, 15(Suppl 2):S1 (15 June 2015)

Care episode retrieval: distributional semantic models for information retrieval in the clinical domain
Hans Moen, Filip Ginter, Erwin Marsi, Laura-Maria Peltonen, Tapio Salakoski, Sanna Salanterä
BMC Medical Informatics and Decision Making 2015, 15(Suppl 2):S2 (15 June 2015)

Using text mining techniques to extract phenotypic information from the PhenoCHF corpus
Noha Alnazzawi, Paul Thompson, Riza Batista-Navarro, Sophia Ananiadou
BMC Medical Informatics and Decision Making 2015, 15(Suppl 2):S3 (15 June 2015)

Detection of sentence boundaries and abbreviations in clinical narratives
Markus Kreuzthaler, Stefan Schulz
BMC Medical Informatics and Decision Making 2015, 15(Suppl 2):S4 (15 June 2015)

SNOMED CT in a language isolate: an algorithm for a semiautomatic translation
Olatz Perez-de-Viñaspre, Maite Oronoz
BMC Medical Informatics and Decision Making 2015, 15(Suppl 2):S5 (15 June 2015)

Exploring Spanish health social media for detecting drug effects
Isabel Segura-Bedmar, Paloma Martínez, Ricardo Revert, Julián Moreno-Schneider
BMC Medical Informatics and Decision Making 2015, 15(Suppl 2):S6 (15 June 2015)

 

Louhi 2014: The Fifth International Workshop on Health Text Mining and Information Analysis

Health Informatics, SYSLAB

The Clinical Text Mining group at DSV organised the 5th Louhi workshop in conjunction with EACL on a beautiful sunny day (April 27) in Gothenburg.

The program consisted of an invited talk by Prof. Sophia Ananiadou from NacTem in Manchester, nine long paper presentations and a poster session.

The workshop proceedings can be found here: http://www.aclweb.org/anthology/W/W14/#1100

Thank you to all participants who contributed to making this day very interesting and memorable!

nigel_maria

Maria Skeppstedt introducing Nigel Collier’s presentation

Poster session with interesting discussions
Poster session with interesting discussions

One minute poster session presentations

One minute poster session presentations

Sunny lunch break

Sunny lunch break

Workshop group photo

Workshop group photo

Two research visits – New York and Boston

Health Informatics, Presentation, Visit

Sumithra Velupillai traveled to the northeastern US coast on February 18-21 for two research visits. In New York, she visited Noemie Elhadad and her group at the Department of Biomedical Informatics, Columbia University, and in Boston, she visited Guergana Savova and her group at the Natural Language Processing Lab, Childrens Hospital Boston and Harvard Medical School. Sumithra gave a talk on both occasions describing her past and ongoing research, as well as current research projects in the Clinical Text Mining group at DSV.

Noemie and Guergana are both part of the ongoing ShARe (Sharing Annotated Resources) project. Data from the ShARe project is used in this year’s ShARe/CLEFeHealth Evaluation Lab 2014’s Task 2 on information extraction (disease/disorder template filling), which Sumithra is co-leading together with Danielle Mowery. Guergana is also one of the main researchers in the Temporal Histories of Your Medical Event (THYME) project, a project very relevant to Sumithra’s current postdoctoral research. During these visits, Sumithra learned a lot from the extensive experience in the area of applying natural language processing technologies to clinical documentation that these two groups have, which hopefully will lead to further future collaborations.

These visits were part of Sumithra’s international postdoctoral fellowship funded by the Swedish Research Council and in part by the Swedish Fulbright Commission.

CLEFeHealth and CLEF 2013

Health Informatics, SYSLAB

Sumithra Velupillai attended the CLEF eHealth lab organized at the CLEF conference in Valencia, 23 – 26 September 2013.

The CLEF conference “promotes the systematic evaluation of information access systems, primarily through experimentation on shared tasks”. This year, there were nine labs with different specific information access problems.

Sumithra was one of the co-organizers of the CLEFeHealth lab together with researchers from Australia, US, Ireland, Finland and more. This year’s lab included three research tasks:

  • Task 1: identification of disorders from clinical reports and mapping of the SNOMED CT disorders to UMLS codes
  • Task 2: mapping abbreviations and acronyms in clinical reports to UMLS codes
  • Task 3: information retrieval to address questions patients may have when reading clinical reports

There was considerable interest in these tasks: 20 countries, 34 teams and 113 submitted systems! During the lab sessions in Valencia, the task teams were invited to describe their work in more detail, the winning teams were invited to describe their solutions along with fruitful discussions with the lab attendees. The lab also contained a student mentoring track for which senior researchers were assigned a mentoring role.

Next year’s CLEFeHealth will include three new information access tasks related to health text.

Two posters at CLEF eHealth, Rome, 17-20 September 2012

Health Informatics, SYSLAB

Mattias Kanhov and Sumithra Velupillai attended and presented two posters at the CLEF eHealth workshop organized at the CLEF conference in Rome, 17 – 20 September 2012.

Mattias presented work about his and his fellow student Xuefeng Feng’s master theses, supervised by prof. Hercules Dalianis, in the poster Natural Language Generation from SNOMED Specifications. Sumithra presented results from the one-year master student Niklas Isenius work on abbreviation detection from Swedish clinical records, co-supervised with Dr. Maria Kvist, in the poster Initial Results in the Development of SCAN: a Swedish Clinical Abbreviation Normalizer.

The conference and workshop was very interesting and Mattias and Sumithra had the opportunity to meet prominent researchers in the area of eHealth and Information Access evaluation from perspectives of multilinguality, multimodality and visual analytics, in the beautiful city of Rome.

Mattias Kanhov with Karl Kreiner from Austria

 

 

 

Clinical NLP workshop, June 21st 2012

Health Informatics, SYSLAB

On June 21st, 2012, we organized a one-day Clinical NLP workshop here at DSV with participants from the Interlock and HEXAnord projects. We had visitors from the US, Australia, Norway, Lithunia, Gothenburg as well as some participants from DSV. The talks were about ongoing research on Clinical NLP in the different research groups. It was a very nice and successful day, thanks to all!

Workshop participants

Hanna Suominen’s presentation and participation over Skype

Mike Conway’s presentation

Danielle Mowery’s presentation

Brian Chapman’s presentation

Wendy Chapman’s presentation

Dimitrios Kokkinakis’ presentation

Maria Kvist’s presentation

Laura Slaughter and Øystein

Hans Moen’s presentation

Dinner cruise to Waxholm in the evening

A beautiful evening

Presentation in Oslo, May 30th

Health Informatics, Presentation, SYSLAB

I was invited to give a talk in a seminar series held at the Language Technology Group (LTG), Department of Informatics, University of Oslo, on May 30th. The talk was entitled Shades of Certainty — Working with Swedish Medical Records and the Stockholm EPR Corpus and was about my recent dissertation work as well as ongoing and planned research within the Health Care Analytics and Modeling research group here at DSV.

It was a wonderful visit, thank you for inviting me, for interesting feedback and questions, and for a fantastic day and evening!

happy audience

13th International Conference on Intelligent Text Processing and Computational Linguistics

Health Informatics, Information Systems, SYSLAB

Sumithra Velupillai attended the 13th International Conference on Intelligent Text Processing and Computational Linguistics (CICLing 2012) at the Indian Institute of Technology Delhi, New Delhi, India, March 11-17.

At CICLing 2012, Sumithra presented the paper Fine–grained Certainty Level Annotations Used for Coarser-grained E-health Scenarios — Certainty Classification of Diagnostic Statements in Swedish Clinical Text, co-written with Dr. Maria Kvist. In this paper, the certainty level annotations in the Stockholm EPR Diagnosis-Factuality Corpus are used for classifying real-world e-health scenarios: adverse event surveillance, decision support alerts and automatic summaries. Promising results are obtained, using simple local context features. Annotation discrepancies are analyzed qualitatively through manual corpus analysis. The main findings are that it is feasible to use a corpus of fine-grained certainty level annotations to build classifiers for coarser-grained real-world scenarios: 0.89, 0.91 and 0.8 F-score (overall average). The paper is published in the Springer LNCS-series.

The conference organized tours to Agra, Delhi and Jaipur, which was very nice!

The 2012 CICLing conference

Sumithra at Taj Mahal